NIRCam Imaging Sensitivity

The exposure times vs. flux estimates at signal-to-noise ratio = 10 in NIRCam images presented here have been obtained using the Pandeia JWST Exposure Time Calculator (ETC) Python engine.

JWST sensitivity update

Preliminary results from observatory commissioning indicate that system throughput is higher than pre-mission expectations, by amounts that vary with wavelength. Until detailed results are available and updated in Pandeia, results calculated by the ETC and JIST tools will underestimate S/N. An announcement will be made when the tools are updated.

On this page

The Pandeia Exposure Time Calculator should be used for all observation planning. This article provides ETC results determined using the Python engine to loop through many calculations of signal-to-noise ratio (SNR) for various readout patterns and exposure times, given the assumptions detailed below. These calculations are then interpolated to determine depth (SNR = 10) vs. exposure time. The script is available on Github.

Depth vs. exposure time

The NIRCam Imaging and NIRCam Sensitivity articles show SNR = 10 sensitivity estimates for imaging in all NIRCam filters given a total exposure time of 10 ks (166.7 minutes = 2.78 hours). Figure 1 shows similar estimates for a range of exposure times in 7 wide filters and one medium filter.

Sensitivity estimates can vary significantly depending on the background and the assumed photometric aperture sizes as discussed below. Please use the Pandeia Exposure Time Calculator to plan your observations.

Figure 1. Depth vs. exposure time

Depth (10-sigma) vs. total exposure time for 4 exposures of NIRCam imaging in 7 wide filters and one medium-band filter, assuming the "1.2 × MinZodi" background defined below. Values are for point sources: short wavelength channel photometric aperture of 0.08" radius with 0.6"–0.99" background sky annulus; long wavelength channel photometric aperture of 0.16" radius with 1.2"–1.98" background sky annulus. Depths are interpolated from results obtained with the Pandeia JWST ETC Python engine. Dashed line extrapolations assume depth in units of flux goes as sqrt(t), or in magnitudes: depth = depth0 + 1.25 * log(t / t0). Note 5-sigma depth estimates are 0.75 mag fainter than the 10-sigma estimates shown here.


JWST's background model varies with the target coordinates (RA, Dec) and time of year. These calculations assume the fiducial "1.2 × MinZodi" (1.2 times the minimum zodiacal light) background at RA = 17:26:44, Dec = -73:19:56 on June 19, 2019, as used in the NIRCam Imaging article. The background model for these observations must be generated using the online ETC GUI and then imported into the Python ETC engine.

Figure 2. Background vs. wavelength

Background vs. wavelength assumed in these calculations, as generated by the online ETC GUI for "1.2 × MinZodi" at (RA = 17:26:44, Dec = -73:19:56) on June 19, 2019.

Exposure time

Words in bold italics are also buttons 
or parameters in GUI tools. Bold 
style represents GUI menus/
panels & data software packages.

Recommended readout patterns and exposure times used for the calculations on this page are described below. They are based on ETC calculations that show these to yield optimal signal-to-noise ratios for a given exposure time. RAPID, BRIGHT2, SHALLOW4, MEDIUM8, and DEEP8 yield high signal-to-noise ratios most efficiently and are preferred to maximize depth. The other readout patterns (BRIGHT1, SHALLOW2, MEDIUM2, DEEP2) may be preferred in some cases, for example, to provide a greater dynamic range (with a shorter first group) to sample bright stars before saturation.
Figure 3. Signal-to-noise ratio vs. exposure time for various readout patterns

Comparison of signal-to-noise ratio for various readout specifications for F200W imaging of an AB mag 29 point source. Each point shows the estimate given 4 exposures, each comprising a single integration that consists of multiple groups (between 2 and 10) of a given readout pattern. This analysis shows that RAPID, BRIGHT2, SHALLOW4, MEDIUM8, and DEEP8 yield higher signal-to-noise ratios than other patterns for a given exposure time. These are plotted as filled circles; the other patterns are plotted as stars. Note BRIGHT2 is restricted to a maximum of 4 groups when reading out the full detectors in both modules.
When reading out the full detectors in both modules, RAPID is limited to 2 groups, and most other patterns are limited to 10 groups (except for DEEP2 and DEEP8, which is limited to 20). The table below gives the full range of recommended readout specifications for exposure times between 21.5 and 1052.2 s. The final two columns assume 4 such exposures.

Note we do not generally recommend integration times greater than 1,000 s. After 1,000 s, the majority of pixels would likely be affected by cosmic rays. See discussion in MIRI Cross-Mode Recommended Strategies.

Table 1. Recommended readout specifications for maximal depth in a given exposure time

Readout patternNGROUPSNINTExposure time (s)NEXPTotal exposure time (s)

Latest updates
    Updated exposure times in Table 1

    Corrected RAPID exposure times in Table 1
Originally published